Agent dispatch failed multiple times for the same room without any deployment or configuration changes

Getting Started
1

On 11 July, we experienced multiple consecutive agent dispatch failures for the same room, despite no code deployments, configuration changes, or infrastructure changes at the time.

The dispatch errors occurred one after another for the same room. The affected session IDs are:

  • RM_r5YwyNyzDVd3
  • RM_W4oJWTX5hfGx
  • RM_GXFw7Ar3eCVx
  • RM_XDQJiUQDgxjT
  • RM_XZD3qCiwpLxV

All of the above sessions belong to the same room.

Unfortunately, observability was not enabled, so I don’t have the corresponding logs, spans, or traces to investigate further from my side but i have saved session report for that room if you wanted in json. the linked session for this room was RM_XDQJiUQDgxjT

Could you help determine what caused these repeated dispatch failures for the same room and whether there were any platform-side issues around that time?

2

Looks like those session IDs all share the same room name, osce-790579-123.

The timeline is on July 10th, looks like between 18:19 and 18:43 UTC and doesn’t align with any incidents.

To me, this looks like a connectivity issue on the client side, I see messages in the logs indicating that the client dropped shortly after connecting. In only one of those sessions, RM_XDQJiUQDgxjT, did they stay connected long enough for the agent to join.

3

Yeah makes sense, i will see if it happens again, i have turned on observability again for few days to see if happens and i could capture anything

4

i havent noticed it again from many days, we can assume closed, might be connectivity issue on the client side.